KEY WORD INDEXING MODELS
Computers began to be used to aid information retrieval system in the 1950s. The first use of computers in information retrieval was the production of indexes. The Central Intelligence Agency (CIA) of the USA is said to be the first organisation to use the machine produced Key Words from Title Index since 1952. H P Luhn and his associates produced permuted title indexes in the International Conference of Scientific Information held at Washington in 1958. Luhn named his index as '"Key-Word-in-Context" (KWIC) index and reported its method of generation in a paper in 1959. The success of KWIC was established after its adoption by American Chemical Society in 1961 in its publication "Chemical Titles". "Keyword" means subject denoting words, chosen mainly from the titles and/or sometimes from abstract or text of the document for the purpose of indexing. The words chosen may be single words, multi-words or even phrases that convey content information of the document. However, the system developed by Luhn was from the words in the title of a document. Several keywords may be chosen for a document to provide access form different approaches of the user. Since the keyword indexing is based on natural language terminology of the documents, this system is also known as "Natural Language Indexing System".
The KWIC index, developed by H P Luhn, is said to be one of the earliest and successful computer-generated keyword indexes. In his method, he suggested the selection of words from the title excepting the unwanted or insignificant words. While the words will form index term, other words in the title, what he said as will be wrapped around it". These words will serve as the context. KWIC indexing system is based on usage of natural language terminology to generate the index entries. All of the words in the titles of a batch of documents for which an index is required are matched, by a computer against a stop-list. This stop-list or stop-wordlist is a record of words which are insignificant in an index. They include words like articles, auxiliary verbs together with such general words as "aspect", "different", "method", "very", etc. Depending upon the subject orientation of each major search system has defined their own list of "stop-words". Some words which might be feasible access points in a general index prove worthless in an index devoted to a special subject area. Indiscriminate marking articles, prepositions, etc. may create problems because of important scientific and technical terms such as "Vitamin A", "On line", etc. In view of this, words to be included in the list of "stop-words" are required to be selected in the light of the subject orientation of the index. Stop-words do not appear as entry words but they are displayed in the titles in the index in order to provide the context of the document. No controlled vocabulary is required for keyword indexing. Indexing terms are selected from the natural language of documents. In addition to KWIC index, there are a number of varieties of keyword indexes that have been developed over the years. Two most important versions are Key Word Out of Context (KWOC) and Key Word And Context (KWAC). They differ only in terms of their formats but indexing principles/techniques remain more or less same.